HARPER@finfun.bitnet ("Robert Harper ", Finland) (06/20/89)
Since there has been some interest in classification programmes recently on this list I thought it would be appropriate to cross post this notice from CLASS-L@SBCCVM. --------------------original----------------------------------- From: "Willem N. Ellis" <A429WILL@HASARA11.BITNET> Subject: Phylip 3.2 Sender: "Classification, clustering, and phylogeny estimation" <CLASS-L@SBCCVM.BITNET> Dear classifier, A few weeks ago Joseph Felsenstein made the source code available of version 3.2 of his Phylip package. Most of these programs have been slightly adapted and compiled by me for working on the Macintosh. The modifications implied among other things the use of the Macintosh on-board random generator and making the programs a bit more interactive. The programs have decidedly not received a really Macintosh-style makeup (contrary, for instance, to the splendid user interface of MacClade). I have not compiled the programs for plotting trees. These are hardware (printer/plotter-) dependent, and are not really useful for Macintosh users, who have a broad range of graphic packages to choose from. Neither have I compiled the "move"-programs. These too are difficult to implement for the Mac, and essentially do what MacClade does at least as good. Perhaps the Maddison's will consider the preparation of a version of MacClade especially directed at molecular data. The programs have been compiled with the Think Pascal compiler. All adaptations have been developed and tested on a Macintosh Plus. Because all 512 Macs around have been upgraded, I couldn't test the programs on a fat Mac. They certainly won't work on a 128. The package seems to work well on a Mac II. The programs have been compiled with the LightSpeed Pascal compiler, which produces output that is reportedly able to make use of a mathematical coprocessor; consequently, the programs must run quite a lot faster on a Macintosh II. Since I do not have such a machine available, I cannot be very definitive yet. A hard disk is handy but not necessary. The compiled (working) programs (together with their documentation etc.) come on two 800 K diskettes. If you wish to receive them, just send me two blank, formatted diskettes, preferably in a reusable envelope (no address label, I am obliged to use university labels). There is no charge. In case you wish to have to original code of the programs as well (both Felsenstein's untouched code and the modified code in the version used by my compiler) you must additionally send me three diskettes. Please do this only if you really will be using this code - the number of customers I have to cater is growing, and copying diskettes is no big fun. Personally, I distribute five more programs of my own hands that do not form a part of the Phylip package, viz TreeTools, BinTree, and Jelly, Shaker, and Cladarea. TreeTools is a general character state Wagner network program; it is slow but enables all sorts of character definitions. BinTree is like TreeTools but accepts only two-state characters and consequently is much faster. Jelly is a modification of TreeTools, specially designed for the character state analysis of allele frequencies; it uses the HAP-algorithm. Shaker and Cladarea are for area cladistical analysis. If you wish a compiled version of these programs (+ short manual) send me one 800 K diskette. Of course, these programs are free as well. The source code of these programs is available from the class-l server. Sincerely, dr. Willem N. Ellis INSTITUUT VOOR TAXONOMISCHE ZOOLOGIE, afd. ENTOMOLOGIE Plantage Middenlaan 64, 1018 DH Amsterdam, The Netherlands tel. (020) 525.6240 e-mail: a429will@hasara11.bitnet ------------------------------------end-------------------------------------